METHOD AND APPARATUS FOR IDENTIFYING AND CLASSIFYING QUERY INTENT

• US 20090077047A1
• Filed: 08/14/2006
• Published: 03/19/2009
• Est. Priority Date: 08/14/2006
• Status: Active Grant

• Current Assignee: Oracle OTC Subsidiary LLC (Oracle Corporation)
• Original Assignee: Inquira Incorporated (Oracle Corporation)
• Inventors: Bierner, Gann Alexander, Beghelli, Filippo Ferdinando Paulo, Cooper, Edwin Riley, Dukes, Michael Peter
• Granted Patent: US 7,747,601 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 40/35 Discourse or dialogue representation
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation

Abstract

Linguistic analysis is used to identify queries that use different natural language formations to request similar information. Common intent categories are identified for the queries requesting similar information. Intent responses can then be provided that are associated with the identified intent categories. An intent management tool can be used for identifying new intent categories, identifying obsolete intent categories, or refining existing intent categories.

Claims

1. A method for identifying query intent, comprising:

using natural language analysis to identify multiple different queries that use different natural language formations to request similar information;

identifying intent categories for all of the identified queries requesting similar information;

formulating natural language words or phrases, or various combinations of natural language words or phrases, for each of the identified intent categories, the formulated natural language words or phrases summarizing in a natural language word formation a common generalized information request intent for all of the different natural language formations associated with the same intent categories; and

associating intent responses with the formulated natural language words or phrases for each of the intent categories that provide common responses applicable to all of the queries in the same intent categories.

2. A method according to claim 1 including:

receiving new queries;

identifying the intent categories associated for the new queries by comparing the natural language formations in the new queries with the formulated natural language phrases summarizing all the different natural language formations associated with the intent categories; and

displaying the intent responses associated with the identified intent categories.

3. The method according to claim 1 including:

providing one or more ontologies that associate different concepts related to a particular industry, the ontologies configured to link a plurality of concepts at multiple different concept domain layers, wherein the concepts associated with the different domain layers are represented by natural language words and the concepts for higher concept domain layers are represented by natural language words having more specific terms than the natural language words representing the concepts associated with the lower domain layers;

identifying the concepts in the ontologies corresponding with elements in the queries; and

using the identified concepts to identify intent categories for the queries.

4. The method according to claim 3 including:

analyzing how the identified concepts and other words in the queries are arranged in sentence structures to identify different natural language meanings; and

identifying the intent categories having natural language words or phrases paraphrasing the identified natural language meanings.

5. The method according to claim 1 including:

logging queries for a certain period of time;

identifying at least some of the most frequently queried intent categories for the logged queries; and

providing pre-query intent responses or links for at least a portion of the identified most frequently queried intent categories.

6. The method according to claim 5 including:

receiving a first group of queries that match the intent categories; and

providing pre-configured intent responses for the first set of received queries corresponding with the matching intent categories.

7. The method according to claim 6 including:

receiving a second group of queries that cannot be matched with the intent categories;

using a natural language engine and one or more ontologies to identify concepts in the second group of non-matching queries; and

using the identified concepts to search for responses associated with the second group of queries.

8. The method according to claim 7 including:

receiving a third group of queries that cannot be matched with the intent categories and cannot be matched with the concepts in the ontologies;

identifying keywords in the third group of queries; and

using the identified keywords to search for responses associated with the third group of queries.

9. The method according to claim 1 including:

receiving queries;

identifying intent categories for the received queries by comparing the natural language phrases representing the intent categories with the natural language phrases in the queries;

providing intent hierarchies for at least some of the identified intent categories wherein the natural language phrases representing the intent categories at higher domain layers of the intent hierarchies have more specific terms than the natural language phrases representing the intent categories at lower domain layers; and

displaying intent responses corresponding with a location of the identified intent categories in the intent hierarchies.

10. The method according to claim 3 including:

receiving queries;

identifying intent categories for the received queries;

assigning parameters to at least some of the identified intent categories that correspond with the concepts in the ontologies; and

displaying intent responses corresponding with the assigned parameters for the identified intent categories.

20. The method according to claim 1 including:

correcting spelling in the different queries;

correcting capitalization and punctuation in the different queries;

performing speech analysis in the queries to identify different types of words in the queries;

performing stem analysis on the queries to add other forms of similar words to the queries;

performing concept analysis on the queries to identify different words that have similar meanings as the words in the queries;

performing linguistic analysis on the queries to identify subject categories for the words in the queries; and

performing intent analysis on the subject categories and concepts identified in the queries to automatically generate the natural language words or phrases summarizing the different natural language formations associated with the queries in the same intent categories.

11. A search engine, comprising:

a processor configured to;

receive different queries and then conduct a linguistic analysis that identifies different concepts and linguistic characteristics for different natural language word formations in the different queries;

identify the received queries having similar information requests according to the identified concepts and linguistic characteristics for multiple different natural language word formations and classify the queries with similar information requests under similar intent categories;

formulate natural language words or phrases or different combinations of natural language words or phrases representing each of the intent categories that use a natural language word formation to paraphrase a common generalized information request for the multiple different natural language word informations used in the received queries that are classified under the same intent categories; and

the processor further configured to provide common pertinent information responses to all of the similar information requests classified under the same intent categories.

12. The search engine according to claim 11 wherein the processor is configured to operate an Intelligent Matching Language (IML) engine and use one or more ontologies to identify the concepts and linguistic formations in the queries and compare the concepts and linguistic formations identified in the queries with concepts and linguistic formations in the natural language words or phrases representing each of the intent categories.

13. The search engine according to claim 12 including a memory that stores preconfigured intent responses for only a portion of most frequently queried intent categories and returns the preconfigured intent responses for the intent categories associated with the received queries.

14. The search engine associated with claim 13 wherein the processor is configured to conduct searches using concepts identified using the IML engine and the ontologies when no intent categories can be identified for the received queries and is further configured to conduct keyword searches when no concepts can be identified in the queries.

15. The search engine according to claim 11 wherein the intent categories are arranged in intent hierarchies and the processor is configured to provide intent responses corresponding to locations of the intent categories in the intent hierarchies.

16. The search engine according to claim 11 wherein the intent categories include different parameters associated with different intent responses.

17. The search engine according to claim 16 wherein the parameters are associated with ontology concepts and the processor displays intent responses corresponding with the ontology concepts.

18. The search engine according to claim 16 wherein at least some of the parameters are associated with user frequencies indicating how frequently users use the search engine, wherein the processor is further configured to provide responses according to their identified user frequencies.

19. The search engine according to claim 16 wherein the intent responses include response parameters associated with additional intent responses and the processor is configured to identify elements in the queries associated with the response parameters and display the associated additional intent responses.